NYC Green Taxi Trip Records for December 2016, Transformed for Regression Benchmarking

December 2016 trip records for New York City's green taxi line, originally provided by the NYC Taxi and Limousine Commission. The dataset was transformed for a tabular regression benchmark, with 'tip_amount' as the target variable and specific columns removed to alter feature importance. It includes only credit card payment trips, as tips are not reliably recorded for other payment types.

Use Cases

• Predicting taxi tip amounts based on trip details and location IDs.
• Benchmarking regression model performance on a dataset with engineered numeric time features.
• Analyzing spatial trip patterns in New York City using the 'PULocationID' and 'DOLocationID' features.
• Studying the relationship between payment type and tipping behavior in taxi services.

Strengths

• Data originates from the official NYC Taxi and Limousine Commission, a primary source for trip records.
• Dataset was specifically curated for a regression benchmark, focusing on a clear predictive task.
• Preprocessing steps, such as filtering to credit card payments, are explicitly documented to clarify data scope.

Limitations

• Row count and file size are unknown, making suitability assessment for large-scale training difficult.
• Column-level documentation is absent; field semantics must be inferred from the original TLC data dictionary after download.
• The dataset is limited to a single month (December 2016) and a specific taxi line, which may not be representative of other time periods or services.

Provenance

Source

New York City Taxi and Limousine Commission (TLC)

Collection Method

Original trip record data was downloaded on 03.11.2018 and subsequently transformed for a benchmark.

Time Range

December 2016

Geography

New York City, USA